[Attempto] Statement id in article

[Changyuan Yu]

David,
  I proposal to add an ID to a statement in the article, because it
can not remotely add, delete or modify a sentence without an ID, using
position will cause race condition when multiple client try to
operator on the same sentence. The 'ID' here is just the
representation of an java object, which has nothing to do with its
actual meaning. So what I proposal is much simpler than you think.
  My proposal is based on the current status of AceWiki implementation
which sentence is independent of context (correct me if I'm wrong), I
can not write something like "'p:Apple1 is a fruit. The fruit is an
apple." in AceWiki. But if the sentence is not context free in
AceWiki(I hope AceWiki will support this in the future, and this make
AceWiki more looks like a regular Wiki), the actual meaning the a
sentence is different in different position of a paragraph or
different article, there is no requirement of the ID(my proposal
version) for a statement, and current implementation is just OK. The
statement is actual in a paragraph and a paragraph is actual in an
article. In such situation, I think "insert" and "delete" should be
the only operators for sentence, other operator can be implemented by
using these two operators.

On Tue, Dec 13, 2011 at 12:25 AM, David Whitten <whitten at netcom.com> wrote:
> I have expressed my ideas below
>
>
> On Fri, Dec 9, 2011 at 7:57 AM, Tobias Kuhn <kuhntobias at gmail.com> wrote:
>>
>> This is an important point. It is more fundamental than it seems at first
>> sight. In order to find a consistent way how to identify sentences, we need
>> to have a clear idea of what a sentence is. Some questions that have to be
>> answered:
>>
>> - When a sentence is changed, does the changed sentence have the same
>> identity as the original one?
>
>
> I don't think the statement "a sentence is changed" is unambiguous.
> It depends on the type of change.
>
> If a sentence in a particular paragraph is replaced in the paragraph with a
> new sentence, the n-th sentence of that paragraph has changed. The sentence
> which once was the n-th sentence is still a sentence, and the sentence which
> now is the n-th sentence is still a sentence.
>
> If the identity of the sentence is solely dependent upon the order of the
> words and the actual words which are expressed, then the identity of the
> sentence has not changed.
>
> If the identity of the sentence is only partially dependent upon the order
> of the words and the actual words which are expressed, then the identity of
> the sentence has changed, because it is no longer the n-th sentence of a
> particular paragraph.
>
> If we liken a paragraph of text to a sequence of mathematical equations, the
> system of the equations can be changed (i.e. the numbers when placed as the
> values of the variables and which makes the system of equations true) if one
> of the equations in the sequence is replaced by another equation that does
> NOT have the "same" relationship between the variables and values in the
> equation.
>
> This analogy is discussing the identity of the paragraph as much as the
> identity of the sentence.
> The sentences in a a paragraph have a relationship to each other.
>
> Similarly, the words in a sentence have a relationship to each other.  Since
> the associated values of the words are dependent upon the location in the
> paragraph where the sentence is found, if the sentence is no longer the n-th
> sentence in the paragraph, then the values of those words cannot be said to
> be the same.  Another aspect is that the values of other words in the
> paragraph may no longer have the same associated values because the sentence
> is no longer the n-th sentence of the paragraph.
>
> There are sentences which have different expressions, much as there are
> mathematical equations which have different expressions, but which establish
> the same associations between words.  These sentences may have an identical
> associations even whilst consisting of different words and a different
> identity.  Perhaps we could say they have identical "meaning" or "impact",
> as they establish the same relationship between the words which they
> consistent of, both within the sentence and the paragraph.
> We might also call these sentences idempotent if each can be the n-th
> sentence of a paragraph with no change to the association of values with
> words.  Although I think an idempotent sentence might alternately be a
> sentence that is not dependent upon its location in a paragraph.
>
>>
>> - When a sentence is moved (within or across articles), does the moved
>> sentence have the same identity as the original one?
>
>
> In my opinion, the moved sentence has the same identity, but the values of
> the words in the sentence and the associated values of other words in the
> paragraph may no longer be the same.
>
>>
>> - Are sentences with the same text identical?
>
>
> Yes.  The values associated with corresponding words in two sentences may
> not be identical.  The position in the paragraph affects the
> "meaning"/"impact" of the sentence.
>
>>
>> ... if they are on the same article?
>
>
> I assume you define an "article" as a sequence of paragraphs. This brings in
> the identity of the article into consideration.  And also might suggest that
> an idempotent paragraph is one which can appear at any position in the
> article, but has the same "meaning"/"impact".
>
>>
>> ... if they are on different articles?
>
>
> The sentence has the same identity.  The impact of the sentence is
> different.
>
>>
>> - When a sentence is copied (not yet implemented) is the new sentence
>> identical to the existing sentence or do we get two sentence instances?
>>
>
> Copying a sentence does not change its identity.  If a copy of a sentence is
> placed in a position in a paragraph (and by extension placed in position in
> an article) then the impact of the sentence in its new position might be
> different from a sentence with the same identity placed elsewhere.
>
> David Whitten
> PS I don't know if this mail message will be successfully sent to the
> mailing list: attempto at lists.ifi.uzh.ch
>
> PPS: Tobias, did we meet each other in early summer 2003 in Zurich at the
> Uni ?  I came by and met several of the Attempto folks then.
>
>
>> I don't have answers to these questions at this point, but it's something
>> that requires careful consideration.
>>
>>
>> Tobias
>>
>>
>>
>>
>> On 09.12.2011 13:21, Changyuan Yu wrote:
>>>
>>> Hi,
>>>   As I understand, now I can only refer to a statement in an article
>>> with its position. The position of a statement will change over time,
>>> so it can not specify an statement in RESTful service (currently I
>>> just use position). I think we should assign an unique id to each
>>> statement in an article.
>>>
>>> Best regards,
>>> Changyuan
>>> _______________________________________________
>>> attempto mailing list
>>> attempto at lists.ifi.uzh.ch
>>> https://lists.ifi.uzh.ch/listinfo/attempto
>>
>> _______________________________________________
>> attempto mailing list
>> attempto at lists.ifi.uzh.ch
>> https://lists.ifi.uzh.ch/listinfo/attempto
>
>
>
> _______________________________________________
> attempto mailing list
> attempto at lists.ifi.uzh.ch
> https://lists.ifi.uzh.ch/listinfo/attempto
>



-- 
Best regards,
Changyuan